“If you can’t measure it, you can’t improve it.”
Peter Drucker was arguably the world’s greatest management consultant. Two of his most famous quotes are highly relevant for traffic management.
The first is the title above. Management begins with measurement. But if you measure wrongly, you will almost certainly manage traffic wrongly. And nowhere is this more evident than in measures of traffic congestion.
The second Drucker quote: Managers do things right. Leaders do the right things. Again – highly relevant to traffic management. For instance: Capable project managers complete highway expansion projects on budget, on time. But strong city leaders question whether highway expansion really is the best response to traffic congestion and ask, is there a better way?
Here is a small example.  Compare this traffic management action: Convert a general traffic lane into a bus lane (done widely in many cities). It sounds like a good idea, right? And results are measurable. Say the bus lane reduces 10 minutes of delay for 20 buses, carrying 1,000 passengers. Total time saved: 10,000 minutes or 167 hours. If this is your measure – it sounds great. But suppose the added bus lane causes 5 added minutes of delay for 800 cars with 900 occupants.
If you measure performance by “delay per passenger”, the bus lane is a good idea – a thousand people save ten minutes, while 900 people lose five minutes.
But if you measure performance by “delay per vehicle”, 20 buses save time, but 800 cars lose time. Bad idea.
According to Todd Litman, “in recent years, experts have developed more accurate and comprehensive congestion evaluation methods, but outdated practices are still widely used, and decision makers are often unaware of the biases in their results.”
Which city is more congested – New York City or Houston, Texas? If you measure “delay per auto” (the decline in average traffic speeds during peak periods), New York City is worse — 74 hours of delay a month vs. 61 hours for Houston. But if you measure by “delay per commuter”, New York City does far better than Houston (37 hours of delay vs. 48 hours), because Houston is far more vehicle-dependent.
Modern traffic management thinks ahead “backwards”. Start with the desired outcome and the measures that define it. Use comprehensive evaluation measures, that reflect congestion costs and take into account the amount people have to drive during peak traffic, lacking other alternatives. [See Box]. Then, work backward to think about how to achieve that outcome. And always, think carefully about how your policies will drive traffic behavior.
= = = = =
Comprehensive Traffic Management Measures
Here is an example of a comprehensive measurement system: Consider how accessible transport services are, and not just travel speed for those who do gain access. Measure the costs of congestion, not just intensity. Measure delays to ALL travelers, not just motorists. Calculate marginal costs created by road users (by increasing traffic density) and just the costs they personally bear (owing to delays). Take into account that optimal traffic speeds may exceed legal limits; congestion costs may reflect compliance rather than traffic jams. Take into account “willingness to pay for speed gains”. A rule of thumb is, 30% – 50% of average wages for individuals, and 30%-50% of total wage, benefits, equipment and product time costs for commercial vehicles. When lanes are value priced, use ‘willingness to pay’ as a guide. Take into account “higher value trips” (e.g. freight or high-occupancy vehicles). Take into account fuel efficiency — when you decrease congestion, speeds rise and fuel efficiency falls; maximum fuel efficiency occurs at 40-50 miles per hour.
= = = = =
Posted by Shlomo Maital – Professor Emeritus, Technion Institute
 Todd Litman. “Smart Congestion Relief”. Victoria, B.C., Canada Transport Policy Institute. April 24, 2018. p. 8